README.md

metadata

library_name: transformers
tags:
  - generated_from_trainer
  - trl
  - sft
base_model: HuggingFaceTB/SmolLM-1.7B-Instruct
datasets: gabrielmbmb/ifeval-trl
model_name: SmolLM-1.7B-Instruct-IFEval
licence: license
model-index:
  - name: SmolLM-1.7B-Instruct-IFEval
    results:
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: IFEval (0-Shot)
          type: HuggingFaceH4/ifeval
          args:
            num_few_shot: 0
        metrics:
          - type: inst_level_strict_acc and prompt_level_strict_acc
            value: 23.06
            name: strict accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=gabrielmbmb/SmolLM-1.7B-Instruct-IFEval
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: BBH (3-Shot)
          type: BBH
          args:
            num_few_shot: 3
        metrics:
          - type: acc_norm
            value: 4.5
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=gabrielmbmb/SmolLM-1.7B-Instruct-IFEval
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: MATH Lvl 5 (4-Shot)
          type: hendrycks/competition_math
          args:
            num_few_shot: 4
        metrics:
          - type: exact_match
            value: 0
            name: exact match
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=gabrielmbmb/SmolLM-1.7B-Instruct-IFEval
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: GPQA (0-shot)
          type: Idavidrein/gpqa
          args:
            num_few_shot: 0
        metrics:
          - type: acc_norm
            value: 0.45
            name: acc_norm
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=gabrielmbmb/SmolLM-1.7B-Instruct-IFEval
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: MuSR (0-shot)
          type: TAUR-Lab/MuSR
          args:
            num_few_shot: 0
        metrics:
          - type: acc_norm
            value: 1.6
            name: acc_norm
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=gabrielmbmb/SmolLM-1.7B-Instruct-IFEval
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: MMLU-PRO (5-shot)
          type: TIGER-Lab/MMLU-Pro
          config: main
          split: test
          args:
            num_few_shot: 5
        metrics:
          - type: acc
            value: 1.73
            name: accuracy
        source:
          url: >-
            https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=gabrielmbmb/SmolLM-1.7B-Instruct-IFEval
          name: Open LLM Leaderboard

Model Card for SmolLM-1.7B-Instruct-IFEval

This model is a fine-tuned version of HuggingFaceTB/SmolLM-1.7B-Instruct on the gabrielmbmb/ifeval-trl dataset. It has been trained using TRL.

Quick start

from transformers import pipeline

question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
generator = pipeline("text-generation", model="gabrielmbmb/SmolLM-1.7B-Instruct-IFEval", device="cuda")
output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
print(output["generated_text"])

Training procedure

This model was trained with SFT.

Framework versions

• TRL: 0.12.0.dev0
• Transformers: 4.45.1
• Pytorch: 2.4.1
• Datasets: 3.0.1
• Tokenizers: 0.20.0

Citations

Cite TRL as:

@misc{vonwerra2022trl,
    title        = {{TRL: Transformer Reinforcement Learning}},
    author       = {Leandro von Werra and Younes Belkada and Lewis Tunstall and Edward Beeching and Tristan Thrush and Nathan Lambert and Shengyi Huang and Kashif Rasul and Quentin Gallouédec},
    year         = 2020,
    journal      = {GitHub repository},
    publisher    = {GitHub},
    howpublished = {\url{https://github.com/huggingface/trl}}
}

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	5.22
IFEval (0-Shot)	23.06
BBH (3-Shot)	4.50
MATH Lvl 5 (4-Shot)	0.00
GPQA (0-shot)	0.45
MuSR (0-shot)	1.60
MMLU-PRO (5-shot)	1.73